DO INSTITUTIONAL AI POLICIES RELATE TO STUDENT AI DEPENDENCY? EVIDENCE FROM INFORMATION TECHNOLOGY STUDENTS

This quantitative descriptive–correlational study examined whether awareness of and compliance with an institutional AI Use Policy are associated with lower dependency on AI tools among Information Technology students. As generative AI tools (e.g., ChatGPT) become common in higher education, overreliance may weaken independent learning skills, yet empirical evidence linking institutional policy to AI dependency remains limited (Ghimire & Edwards, 2024). Data were collected from 168 undergraduate IT students in the Philippines. The survey measured AI use frequency and purposes, AI dependency using the AI Dependency Index (AIDI; 15 items), study habits, academic performance (GWA), and policy awareness and compliance. Descriptive statistics, t-tests, and multiple regression were applied. Most students reported frequent AI use (73.2% daily or several times weekly), primarily for understanding concepts (83.3%) and coding/debugging (62.5%), with 36.9% using AI for research/summarizing. Overall AI dependency was moderate (M = 2.78). Students aware of the policy reported lower AI dependency than those unaware, t(166) = −2.25, p = .026. In regression, policy compliance significantly predicted lower AI dependency (β = −0.290, p < .001), while study habits (p = .246) and academic performance (p = .168) were not significant predictors. Findings suggest that stronger policy communication and reinforcement may help reduce student AI dependency. Keywords: AI Dependency; Institutional Policy; ChatGPT; Self-Regulated Learning; Information Technology Education; AIDI